Re: [tor-relays] EOMA68 as a platform for trustworthy computing? / was: Exploiting firmware

[Roman Mamedov]

On Thu, 8 Dec 2016 12:11:48 +0100
Christian Pietsch  wrote:

> On Thu, Dec 08, 2016 at 10:41:46AM +0500, Roman Mamedov wrote:
> > On AMD that's been implemented only after "Family 15h"
> > https://libreboot.org/faq/#amdbastards
> > https://en.wikipedia.org/wiki/List_of_AMD_CPU_microarchitectures
> > 
> > Family 15h itself is safe.
> > 
> > It includes FX-series 8-core CPUs at up to 5 GHz supporting DDR3-2133 RAM:
> > https://en.wikipedia.org/wiki/Piledriver_%28microarchitecture%29
> > 
> > So don't handwave-away AMD with "they are doing that too", today you CAN 
> > have
> > a non-backdoored modern high-performance CPU -- from AMD.
> 
> Interesting, but this microarchitecture entered the market in 2012 and
> is probably being phased out now, I guess?

It is still their current desktop/server performance offer today, the next
generation ("AMD Zen", also making the switch to DDR4, and likely adding that
PSP) is planned to be released some time in 2017 and is mostly at the stage of
announcements and rumors right now.

-- 
With respect,
Roman
___
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

Re: [tor-relays] EOMA68 as a platform for trustworthy computing? / was: Exploiting firmware

[Alexander Dietrich]

Can you please move these discussions to a more appropriate mailing list 
(i.e. tor-talk or maybe tor-dev)?


Thank you,
Alexander
---
PGP Key: https://dietrich.cx/pgp | 0x52FA4EE1722D54EB

On 2016-12-08 12:11, Christian Pietsch wrote:

On Thu, Dec 08, 2016 at 10:41:46AM +0500, Roman Mamedov wrote:

On AMD that's been implemented only after "Family 15h"
https://libreboot.org/faq/#amdbastards
https://en.wikipedia.org/wiki/List_of_AMD_CPU_microarchitectures

Family 15h itself is safe.

It includes FX-series 8-core CPUs at up to 5 GHz supporting DDR3-2133 
RAM:

https://en.wikipedia.org/wiki/Piledriver_%28microarchitecture%29

So don't handwave-away AMD with "they are doing that too", today you 
CAN have

a non-backdoored modern high-performance CPU -- from AMD.


Interesting, but this microarchitecture entered the market in 2012 and
is probably being phased out now, I guess?

As far as modular, open hardware is concerned, I set my hopes on the
new EOMA68 architecture. This is a draft standard for computer cards
in the PCMCIA form factor:
http://elinux.org/Embedded_Open_Modular_Architecture/EOMA68

Crowdfunding for the first implementation of this specification
(EOMA68-A20) has been successful:
https://www.crowdsupply.com/eoma68/micro-desktop
I ordered one, and I convinced Digitalcourage e.V. to order three more
that will be used for trustworthy computing – maybe Tor relays.

The performance of this first implementation will probably be
comparable to the CubieTruck which is based on the same SoC, the
Allwinner A20, which is an ARM Cortex-A8 design. So CPU-wise, it will
be between the Raspberry Pi 2 and 3. Network throughput might be
better because it does not use Raspi's peculiar design.

New implementations with more powerful hardware are in preparation
(see the latest updates on CrowdSupply), but we are still talking
about smartphone-like systems.

Last time I checked, Tor did not support the hardware AES acceleration
of the A20 SoC called Security System (SS) .
Is this still the case?

Greetings from Bielefeld, Germany!
Christian


___
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

___
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

[tor-relays] EOMA68 as a platform for trustworthy computing? / was: Exploiting firmware

[Christian Pietsch]

On Thu, Dec 08, 2016 at 10:41:46AM +0500, Roman Mamedov wrote:
> On AMD that's been implemented only after "Family 15h"
> https://libreboot.org/faq/#amdbastards
> https://en.wikipedia.org/wiki/List_of_AMD_CPU_microarchitectures
> 
> Family 15h itself is safe.
> 
> It includes FX-series 8-core CPUs at up to 5 GHz supporting DDR3-2133 RAM:
> https://en.wikipedia.org/wiki/Piledriver_%28microarchitecture%29
> 
> So don't handwave-away AMD with "they are doing that too", today you CAN have
> a non-backdoored modern high-performance CPU -- from AMD.

Interesting, but this microarchitecture entered the market in 2012 and
is probably being phased out now, I guess?

As far as modular, open hardware is concerned, I set my hopes on the
new EOMA68 architecture. This is a draft standard for computer cards
in the PCMCIA form factor: 
http://elinux.org/Embedded_Open_Modular_Architecture/EOMA68

Crowdfunding for the first implementation of this specification
(EOMA68-A20) has been successful: 
https://www.crowdsupply.com/eoma68/micro-desktop
I ordered one, and I convinced Digitalcourage e.V. to order three more
that will be used for trustworthy computing – maybe Tor relays.

The performance of this first implementation will probably be
comparable to the CubieTruck which is based on the same SoC, the
Allwinner A20, which is an ARM Cortex-A8 design. So CPU-wise, it will
be between the Raspberry Pi 2 and 3. Network throughput might be
better because it does not use Raspi's peculiar design.

New implementations with more powerful hardware are in preparation
(see the latest updates on CrowdSupply), but we are still talking
about smartphone-like systems.

Last time I checked, Tor did not support the hardware AES acceleration
of the A20 SoC called Security System (SS) .
Is this still the case?

Greetings from Bielefeld, Germany!
Christian


-- 
  Christian Pietsch · volunteering for Digitalcourage e.V., Bielefeld, Germany
  https://digitalcourage.de | https://bigbrotherawards.de

Vorratsdatenspeicherung? Nicht schon wieder! Unterstützt unsere 
Verfassungsbeschwerde: https://digitalcourage.de/weg-mit-vds

Rettet den Datenschutz – der aktuelle Entwurf für ein neues Datenschutzgesetz 
ist eine Katastrophe: https://digitalcourage.de/rettet-den-datenschutz


signature.asc
Description: PGP signature
___
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays